Giovanni Palla

wtf how can 3.9 and 3.10 not be consistent?

mmh but this is a bit suspicious, like the results seems to put labels with different orders, despite everything being generated by RNG. In the last commit, I took artefacts...

quick push to try #699 where tiling is vectorized, removed the need for `pandas.DataFrame.apply`. Quite big speedup 

that is correct! Thank you @schae211 for this, would really appreciate if you could submit a PR clarifying this line in the docs, would be super useful!

hi @WT215 I would check the output of the segmentation layer first, in order to see how many segmentations masks you have obtained, and whether they are of high quality

I tried with the code above Details ```python from ott.geometry.pointcloud import PointCloud from ott.problems.quadratic.quadratic_problem import QuadraticProblem from ott.solvers.quadratic.gromov_wasserstein_lr import LRGromovWasserstein import jax from ott import utils import numpy as np...

hi @wangjiawen2013 that's parameter is used when building visium neighbor graph, when there is some sway. This is because the actual coordinates of the spots might have some errors, and...

hi @ArneDefauw , sorry for late reply. I took a look at > [gletort/NeubiasPasteur2023_AdvancedCellPose@main/distributed_segmentation.py?rgh-link-date=2024-07-08T08%3A01%3A09Z](https://github.com/gletort/NeubiasPasteur2023_AdvancedCellPose/blob/main/distributed_segmentation.py?rgh-link-date=2024-08-26T09%3A53%3A28Z) it's a nice approach I think, and I wonder if it should be included already here....

hi @LLehner I was just wondering what's the status of this? I see a lot of conflicts and tests still missing. No rush, just curious if there is a timeline

also on this question > How should multiple slides be dealt with? All these methods should work with multiple slides, but i'm still not sure how the adjacency matrix changes...